Iron Man Mk6 (or Mk4) forearm help - How do you do the mini rockets?

Eh... some steps forward, some steps backwards.

The good:

• The cover going front-to-back makes it significantly sturdier. The older prototypes I could flex it and push the lid in, getting it stuck under the ledge it was suppose to rest on. This one is solid.
• The gap between the lid and the cover is very tight. I can barely get a fingernail in on both sides of it. It was slightly tighter, but I was trying to fix some minor snagging so I trimmed plastic off pretty much every joint. (Side note: to keep that gap in the final product I won't be able to paint the sides of the lid or cover - which means I'll need to print them out of either silver or red)
• The rocket size is much better. It's about as big as I can go without completely redesigning how I print them. (Fun fact, they already have cut outs in the back, so they can fit around the moving parts. This gives them the illusion that they're longer than they really are)
• The thicker hinges haven't shown any signs of cracking. Maybe I'm lucky, or maybe I finally have them just thick enough.

The bad:

• The servo linkage is too hit-or-miss right now. Previous iteration barely closed all the way, this one doesn't. Need a better hole position, possibly some sort of slot that I fill in with glue to accommodate for minor variation in linkage arm sizes.
• The holes that I didn't re-enforce did break. Most troubling, the hole holding the servo broke clean in half. While writing this up I realized that the back side is also broken, splitting the base into two parts.The connections between the lid and it's hinges broke as well.
• Servo compartment was just a hair too tight. Not sure how much of this was due to swapping servos for this test, but regardless it should be fixed.
• Cover connections need a lot of fine tuning. Too tight, caused a lot of broken parts before I just hot glued them in place.
• Rockets are only held on by hot glue. I tried printing a pin for them to hold onto, but that didn't work (pin snapped when trying to trim it down to size). Still the shape of the rocket and the platform connection does a decent job of lining the two up. The "fix" might be just removing the pin and hole all together.
• The rocket nosecone just looks bad.... need a cosmetic redesign.

The untested:

• Added little loops at the front and back of the base. The idea is to eventually connect these loops to two rings, which will preserve the spacing and alignment between the rockets. Since the forearm curves a bit it might be a better fit if the rockets bend forward (or backwards) slightly, and that could be achieved by having a different sized ring at the front or the back.
• Didn't get around to testing my servo controller. It didn't make much sense to test it with only one servo running, since it would have just added more wires and troubleshooting into the mix.

So yeah, lots of tweaking to do.



Side note: It has crossed my mind that you could flip the mechanism around and use it for different costume ideas.
For example, if you wanted to make a Pacific Rim style rocket-punch, you could flip the design around and swap out the missiles attachment for a booster rocket attachment.

Some comparison pictures between version 2 and version 3:

View attachment 836234
Gives a sense of just how much bigger the missiles are now than they were in my first test. There's not really much room to make them any bigger (or to lay them flatter, for that matter).


View attachment 836235

View attachment 836236

The top down view shows what changed with the cover. Before it only covered the servo, now it wraps around the mechanism. Interesting to note that you can visibly see the difference in the gap between the lid and the cover, even without applying pressure. The old model has a much more noticeable gap.

(Imgur links, in case the attachment option isn't working)

is it possible to fit a small spring in there to help assist with the closing of the door? Just a thought as you wouldn't need much to help pull the door closed. Just need to find a spring that wont over power the servo arm.

IronDan27 said:

is it possible to fit a small spring in there to help assist with the closing of the door? Just a thought as you wouldn't need much to help pull the door closed. Just need to find a spring that wont over power the servo arm.

Click to expand...

It's not a question of power, the servos I'm using have more than enough power for this task. The trick is applying that power in the right direction. While I got v2 to work, it initially behaved the same way v3 is behaving. The difference was that the v2 design was a hair closer to having the right angle to allow it to be pushed down, so when I got lucky with the hot glue it worked the second time I tried it. If I'd hot glued it at a higher position, then it would have failed again.

A spring would likely introduce more problems than it would solve at this point. It'd make it harder for the servo to open and stay open (which would in turn make all the servos whine when posing for photos - not ideal), and it'd put additional stress on some very small parts - many of which have had to be replaced because they were simply too thin. I'd have to spend a lot of effort trying to find the right spring for the job, and even then it'd have drawbacks.

I think at this point it'd be better for me to look into modifying how I apply the force from the servo to the platform. I want to try lowering the pin hole in the main platform to as low as possible and converting it into a slot. The slot would accommodate slight changes in control arm length, and the lower position should make it easier for the servo to push the platform down. The variance in the length (and time moving in the slot) shouldn't result in a meaningful difference in the time to open, since we're looking at maybe 2 or 3mm at most.

If that doesn't work, then I can try adding something similar to an RC plane's Control Horn to the front platform arm. Depending on how that design goes it might make the overall profile flatter but longer, since the servo arm might not need to travel a full 90 degrees in order to fully open/close.

-----

TL;DR: I've got some ideas on how to fix this and make it work. This version just provided more learning opportunities than I'd expected

Adding a small, light weight tension spring to the center lever portion may help to pull the door to home position.
As a quick test, you could substitute a small rubber band cut to size, to test the concept.



Just thinking out loud.

A rubber band would work too, but either way great design keep up the great work!

I'll keep the rubber band idea in mind as a backup, but it would still make the servos whine.

... Actually, I just had another backup idea from @propmaster2000 's design thread. A loop of string connected to pullys could do the trick. My problem is applying force in the right direction, but if the pullys loop around and up to the cover on the opposite side of the servo then as the servo tried to lower the platform it could pull​ down instead of push down.

I got my changes printed off last night, and they look promising, but haven't had time to fully assemble them yet. I tried printing the cover in a way that would use less support material, but the lid isn't laying flat because of it. :unsure oh well, maybe I can design some custom supports to get around that issue for the next test. If not I can fall back to the method I used before, it just seemed wasteful.

Also, while looking for reference material I found this amazing model someone put together. They took the time to animate everything, including how the larger rocket would interact with the smaller rockets internally.

If I'd been aware of this video at the start of my project I may have designed some things differently, but I'm still pretty happy with what I've got.

While impressive, some of their mechanics are a bit too impracticable to implement 1-to-1, particularly the top rocket setup. My final version will simplify the design by replacing the top mini rocket launcher with the large one. Yeah, I'll technically only have 7 of 8 mini rockets, but I think that's an acceptable trade off for a wider weapon variety. I do think I'll use their video as a reference for final orientation/placement of the rockets and how they line up with the panels on the top.

With videos like this of CGI mechanics, it does give one another way to look at things.

Since most of the mechanics shown is most likely protrayed as being metal (which can take more abuse then
plastic parts of the same size), it can be shown as small, miniature parts crammed in to a compact area and
still leaving room for the person inside.
In the world outside of CGI, it is just not possible especially if the parts are 3D printed PLA plastic.

I did notice in one of my video shot (captures), that there didn't seem to be anything holding the door from pivoting on it's
own axis (at the lifting mech. arm area).

The arm attachments look to be a pivot point at the door and nothing holding the rear..




P.S. DrCyanide - Looking forward to your next video post.

.

.

propmaster2000 said:

With videos like this of CGI mechanics, it does give one another way to look at things.

Since most of the mechanics shown is most likely protrayed as being metal (which can take more abuse then
plastic parts of the same size), it can be shown as small, miniature parts crammed in to a compact area and
still leaving room for the person inside.
In the world outside of CGI, it is just not possible especially if the parts are 3D printed PLA plastic.

Click to expand...

A combination of precision metal work and something far more compact and refined than my 9g servos would be needed to bring it to life like that, absolutely. Their CGI hints at miniature pistons, something that's a pretty consistent theme in the movies. Pneumatic or hydraulic... pneumatic for pieces like that which don't have to do heavy lifting is probably more canonical, considering the Mk 42's arm didn't gush hydraulic fluid when he had to rip the missile out of it. Hard to imagine the Mk 3 lifting a car with only pneumatics, but then again that was a struggle.

I digress. Regardless, it's a fun concept piece, even if some of the mechanics would be easier to make scaled up.

propmaster2000 said:

I did notice in one of my video shot (captures), that there didn't seem to be anything holding the door from pivoting on it's
own axis (at the lifting mech. arm area).

The arm attachments look to be a pivot point at the door and nothing holding the rear..

View attachment 837070

Click to expand...

Actually, that one doesn't bother me. Yeah, they added a slight twist there in the CGI, maybe 15 degrees, but I actually get remarkably similar movement with my own lid and it hinges just like that. The third connection (the one you suggested a rubber band be connected to) is only there to make it easier to move the lid at the same time as the rockets, it's not there for stability.

We'll see if that third arm ends up sticking around in my design. The more I played with the v3 print the more I realized the third arm actually shifts a moderate amount of force back down the base near the servo. The crack/split at the servo mount may have been in part from that force, only showing up with the v3 print because the v1 and v2 had more flexibility at the top without the cover. The v4 base is reinforced in that area, but if it continues to be a problem I may need to look into another method to raise the lid.

propmaster2000 said:

P.S. DrCyanide - Looking forward to your next video post.

Click to expand...

Me too.

I see looking at your design, you have more brackets going to your lid to stableize it then shown in the CGI version image which I attached earlier.
In the attached image, you don't see anything holding the CGI version lid (door) stable.

Keep up the good work.
Post updates when you can.

.

Alright, here we go again.


After making the video, I removed a piece of hot glue I had added in an earlier attempt to replicate the v2. With that glue removed it basically lays flush when closed, only sticking up due to flaws in the print itself. I'm feeling too lazy to re-record and upload though.

The Good

• It closes!
• It still opens. (That shouldn't be taken for granted, you know!)
• There's some slack for control rods to be slightly different lengths.
• The rounded missiles look better.
• The cover fits correctly.
• The filament pins work very well! I've taken this version apart a lot of times trying to test various things, and they're still holding nicely.

The Bad

• There's something funky going on with the closing mechanism. (More on that below)

The Ugly

• The cover's inset groove (where the lid should rest) didn't form right due to bad supports or completely missing supports. I need to look into making custom supports to prevent that.
• The rockets have their cut-outs based on the v3 control arm holder. That should be redesigned, potentially including a detail that looks like it's holding the rocket.
• Some walls look a little thin and prone to break. Not a model issue, just a print settings issue.

So yeah, it's nearly there! The last major thing is to fix the closing mechanism.

When it closes right now, the hinges and the back of the rockets smash into one another. After a moment of this, the two push past each other and the platform "snaps" into it's closed position. What drove me nearly mad over the last two days was that it doesn't close properly without the rockets installed, and since I was treating them like the decoration they are suppose to be I wasn't getting the closing motion I was expecting. Instead of closing all the way, it'd get most of the way there, then the platform would rotate into a nose-dive like stance. Since the back of the platform is what's connected to the lid, the lid wouldn't get pulled down all the way. I removed a lot of plastic that I thought might be causing the issue, but none of that was fixing it.

I won't be satisfied with the design until I can fix this closing bug, but at least there's another working version now! Hopefully between the v2 and the v4 I can find out what works, what doesn't, and make it more consistent. After it's more consistent I can work on printing a full test ring and making it look more like the forearm.

Now that I've thought it over, I think I know why both versions worked out.

v2
The angle that I hot glued the control rod at happened to line up nearly perfectly with the angle between my filament hinges and the center of the servo. This basically made the control rod like an extension of the platform, and the servo arm like a third arm moving the platform up and down. That's why angles didn't matter very much on this version - because it was the arm doing all the lifting.

v4
Worked because the hinge at the back stopped the platform from rotating like it wanted to. If something else could stop the rotation, it would work without needing to bump into the rockets.

v5 plan
I'm going to add a third arm to the platform, opposite the control rod. The third arm should stabilize the platform and prevent it from rotating. I'm going this route instead of the v2 route because it will be easier to design and assemble, and wouldn't get messed up if I decide to increase/decrease the length of the arms at a later date.

It seems like you have a few options where some work and some not so much.
It is almost like a "snap action" switch set up now when it closes.

Maybe, there is a way to add a second lift and push shaft from the servo arm to the door bracket as well?
So the servo to doing two things at once.....

Keep up the good work.

.

Just checking back for updates (if any)

.

I was just fiddling with the latest version in hand as you posted that. It's not quite working as I expect it to, and I'm trying to figure out if it's because of some problem with the print, the design, or some other factor which I hadn't considered yet.

I added the third arm. It seems to be doing exactly what I intended it to do: preventing the platform from going into a nose dive position when being pushed down. It has a thinner hinge connection than the other arms, but it seems to be holding up well enough that I might redesign the whole platform to use that method and make it thinner/lighter.

The trouble is in that the platform seems reluctant to go all the way down when operated by the servo. At first I thought it was something to do with the lid and cover not lining up properly, but I've removed them both from the t est version and it's still having issues. To make it even more puzzling, when I rotate the servo arm by hand most of the time it closes just fine.

Very annoying being 2-4mm away from the main mechanism being complete. Just close already and let me move on to the next problem! :behave

I'm going to tinker with it a bit before getting a video together. Need some time to figure out what's causing this latest annoyance. I saw some minor improvement by changing what angle the servo was trying to hit, but mostly it was a lot of servo whining with very little change.

Beyond that, I've taken some thought to how I'll eventually have to maintain this in a costume. I figure that the cover and lid would be painted all fancy, and if it's to be a gapless loop between the individual sections then none of the covers can move relative to each other - or even rotate - otherwise they'll break the paint job of the final project. The old way of connecting the cover to the base wouldn't do, since you'd have to paint the whole arm from scratch every time you needed to replace a burnt out servo or fix a control arm that got loose. So with this version I've connected the mechanisms to the cover via some pins in the front and back. It's a decent first pass, but it'll need some tweaking as well.

... And I might have just figured out why it was all struggling to work.

When designing this, I included some wiggle room for the parts. The hinges are 0.2 mm wider than the arms that go in them. I roughly positioned the arms, then went on to design the next hinge.

What I didn't do was make sure that those new hinges were lined up the same as the ones on the other end of the arm. This resulted in an ever so slight error, where the hinges on the base weren't in line with each other, and the hinge on the lid wasn't in line with the hinge on the platform. Across those two joints I was getting more friction and a twisting motion that the third arm couldn't possibly negate. All because I wasn't careful enough when designing it in the first place. :facepalm It may also explain why I was constantly having to carve away at pieces to get them to slide past each other nicely (although some of that is just due to how printers handle tight corners).

I'm going to take a moment to double check my updated design, then print it out. Since this effects every major part, I think I'll save a video until I get this new version assembled.

OK... This update is less glamour shots, more trouble shooting.

It looks like there's an issue with the mechanism that connects the lid to the platform. It's just requiring too much force to get past consistently. Honestly I had some misgivings about that link from the first design, but I kept with it because it looked like it was decently functional when moving everything by hand. I do blame that linkage for some breaks in other parts of the design though.

I think it's time to look at other designs to perform the same function, possibly even the springs/rubber bands which have been suggested earlier.

I'll think on it a bit tonight and the next few days and see what I can come up with. I'm too tired to throw down ideas right now.

I'd hoped to get the mechanics ironed out by now and start messing with the other goodies I've got planned. Ah well, all in good time.

\/

So far so good.
Maybe you could disguise the spring to look like a hydraulic piston?
One small tube inside a larger one with the spring inside......

Looking forward to your next improvement.

.