WIP Apollo Flight Director Attitude Indicator (FDAI) "8-Ball"

[HerculesFerrari]

I recently came across the following blog post about an attitude indicator from an F-4 Phantom and pondered to myself how cool it would be to have an Apollo Flight Director Attitude Indicator (FDAI) to go with my replica DSKY. Having found no suitable models in any of my searches I went into research mode and set about doing it myself. Working from Block II reference photos, flight journals, manuals, and mil specs, I modeled it in Fusion. I'm uncertain if the originals were technically etched vs. engraved but the markings were very deep and sharp so I've been treating them as engravings. As such, all lines, tick marks, and numerals need to have the same uniform width and depth. Could they simply be debossed and get close? Sure, but I'm not looking for just close and debossing results in distortions on the numerals because it assumes linear projections onto the sphere. That meant I needed to create engraving line models for the numerals based on the Futura Medium font in order to get perfect alignment and a uniform engraved appearance. The trick there was to center each numeral over a sphere, creating a surface going back to the origin, trimming it to the sphere, and using that to create a 3-d sketch path along which to sweep a 1/64" sphere.

After sending the two half sphere models to my resin printer, and knowing I'd need to sand it smooth anyways, I filled the lines with acrylic paint markers and went to town. Trying to work with half spheres was clumsy at best so I modeled a little ring fixture to fit inside the shells and keep them aligned so I couldn't screw up by sanding and rounding over the equator.

At this point it has turned out so pretty that I which I didn't need to strip it and start spraying, but here we are.

Comparison with the FDAI on Apollo 15:

 

[HerculesFerrari]

Update:

After reviewing scan data from Smithsonian and many photos, I'm confident this first design is undersized at ∅3.5" which also fails to meet the minimum ∅3.75" in the mil spec. Images consistently show ~90° of the ball are visible with a 3" window. A quick bit of trig shows that 3"/sin(45°) yields ∅4.25" which should still have plenty of room inside the 7" enclosure.

This is all just as well since I need to print off a new one to incorporate a 3-axis gimbal anyways. The new design is ~95% complete, with only placement of numerals remaining before I can run the print(s?), probably two prints to be safe. I'll do most of the gimbal mechanism in FDM. Should have some cool stuff to show off in a few days in any case.

In other news, I found a great price on a surplus ARU-11A. The original data plate is missing but I noticed a Lear Siegler part number engraved on the edge of the bezel. While it's not an exact match for an Apollo FDAI it should be *very* similar and I'm sure I'll learn some things having it on my bench. There's a good chance I'll work up a CAD model from it that's worth sharing for the interested. As an aside, Lear *did* make ADIs for the Shuttle program used in simulators (flight hardware was all Honeywell afaik). Does that mean I might need to work up a Shuttle ADI? ... probably, I doubt I can resist.

 

[HerculesFerrari]

First half just finished printing, which leaves the other hemisphere and equator (which I did run off on FDM to verify I would be happy with the bearings I have sitting around. I figured I would share an exploded view of what will be going on: the equator will have a central bearing to which a shaft from each hemisphere will attach and screw together so they are coupled and can rotate freely.

The colors represent my current best information on the federal standard colors that would have been used on the originals (I'll share more on that when I'm more certain of their authenticity).

The gimbal itself will be pretty straightforward and then much less so as motors and slip rings are added.

 

[The Kaiser]

This is super cool and looks fantastic! I picked up a large lot of vintage Master Specialties Twist-Lite switches a couple year ago with the intention of fully replicating one of the Apollo-era flight consoles. Would you be willing to share the file you made for this print? I would love to see additional photos as this project progresses. Thank you for sharing.

 

[HerculesFerrari]

I'll certainly be sharing progress and appreciate the kind words. None of the design work is polished enough yet but I will make it available in the future (probably from my Cults3d page) . From my own experience it's starting to look like it can turn out great as a pure resin print for static display, but to look its best and/or be capable of spinning, a metal equator disc will be nearly essential.

Scale up matches the known bezel dimensions as I expected (verified here with a quick and dirty print). The ARU-11A arrived the other day and is already helping me to understand the bezel, lights, needles, pointers, and mechanism immensely. It may be a scary close match to what was used on the shuttle, and was made by the same contractor, so I can promise a corresponding ball for that as well.

As for your project, I'm prototyping duckbill switch tips and they're almost ready to go. They look good in FDM, pictured below) and I only need to verify in resin before I share them to go with my switchguards. On that note, I have a couple of Apollo switchguard models I never got around to sharing and will do so soon.

Thanks again for following along!

 

[HerculesFerrari]

For a bit over $100 and missing its manufacturing labels I was expecting a paperweight with a gimbal. This thing looks complete and I've found several parts in common with the attitude indicator used on the Space Shuttle (also made by Lear Siegler). I'm sure I'll learn a lot from it but I can't cannibalize it. Vintage electromechanical aerospace stuff is engineering artwork.

Consider it a future project to get all the synchros running.

 

[AmazonChitlin]

For a bit over $100 and missing its manufacturing labels I was expecting a paperweight with a gimbal. This thing looks complete and I've found several parts in common with the attitude indicator used on the Space Shuttle (also made by Lear Siegler). I'm sure I'll learn a lot from it but I can't cannibalize it. Vintage electromechanical aerospace stuff is engineering artwork.

Consider it a future project to get all the synchros running.
View attachment 1876072View attachment 1876073View attachment 1876075

I’d be interested to see if you put power to it if it’s gyros would spin up.

Very cool piece of hardware. When I learned to fly one of the planes I’d fly fairly regularly had attitude indicator kind of like this one (color wise). It was always a little more confusing at a glance than the blue/brown modern ones. I feel like it’d be even more confusing with it being a combination ADI, and compass! I do believe these tumbled fairly regularly as a result of the complex movement. I could be wrong though

 

[HerculesFerrari]

I’d be interested to see if you put power to it if it’s gyros would spin up.

Very cool piece of hardware. When I learned to fly one of the planes I’d fly fairly regularly had attitude indicator kind of like this one (color wise). It was always a little more confusing at a glance than the blue/brown modern ones. I feel like it’d be even more confusing with it being a combination ADI, and compass! I do believe these tumbled fairly regularly as a result of the complex movement. I could be wrong though

I'm working on testing it out. There's no chance of spinning up though because it requires external signals from an IMU. As a 3-axis ADI it's a very close relative of space-flown hardware and would experience only limited drift. It's truly a fascinating beast.

Fun challenges include sourcing a 115v 400Hz power supply and seeing if I can synthesize phase shifts to control synchro position.

 

[HerculesFerrari]

After finding a phenomenally clear image from the Apollo 11 flight journal (I'll be sharing it soon) and getting some help from a couple of real FDAI owners I have the CM 8-ball mapped out and modeled better than ever. There were two line widths ~0.010" and 0.020" that are now present, I am updating all numerals and am quite confident they were spec'd to be 3° tall, the list goes on. I'm excited!

While researching this topic I came across an old auction for a LM FDAI and the mfg data plate actually names a Grumman Specification Control Drawing (I've never seen one on a label, so it's a potentially great find). I'll be reaching out to NARA Ft Worth to see what I can see. Every indication is that the LM units by Lear Siegler went through some minor modifications and were adopted into the Shuttle program.

I'll be back with pictures soon and more updates.

 

[HerculesFerrari]

Refining meshes can decrease time spent sanding dramatically. I used a layer height of 0.03mm as before and it came out so clean I was able to dive in at 400grit, and this is just a 4k printer with 0.05mm pixels. The picture is before sanding. Adjustments made from the bottom reference photo from Apollo 11's flight journal were mostly related to line width with major lines, all 30° intervals, increased to 0.02" while others were decreased to 0.01". Line depth was increased to 0.02" from 0.0075" used in all previous prints. I don't know if any competent engraver would actually do that to themselves but it sure wicks up paint and looks more like the real things this way.

Since this print I've been reading more relevant mil specs including MIL-M-18012B MARKINGS FOR AIRCREW STATION DISPLAYS DESIGN AND CONFIGURATION OF. Guess who has two thumbs and copied the wrong font. In fairness it's extremely similar to Futura Medium but the correct options for engraving were Gorton Extended and Gorton Moderne, and Gorton Moderne is a dead ringer. I'll use their typeset catalog as a reference for now and start lurking around typography circles to see if I can find anything better, but I'm sure I'll get close enough either way.

Who knew replicating a sphere that's marked up like a ruler could get this involved? I didn't even know that 3d electrohydraulic hand guided pantographs existed, yet that's almost certainly how numbers were engraved on a sphere back then. Searching for how it would've been done pre-CNC mostly brings up CAM forums and CAD tutorials. Old machining catalogs make better sense of it.

Time to go make a fresh shiny set of numbers.

 

[HerculesFerrari]

Screenshot of the single line Gorton Moderne numerals I've whipped up (if anyone wants a copy of them I have them in dxf ). I'm trying to source a master copy type set and will put that on GitHub whenever it comes together.

New hemispheres are coming off the printer incorporating the correct typeset and a minor change to some tick mark lengths. I'm probably splitting hairs since the biggest change was still less than 0.010" but what's the fun if you aren't keeping the rivet counters happy, right?

I may need to look into an 8k printer because there are some subtle things that aren't coming together yet. Then again I may need to have a master version properly done via rotary engraving with a vintage 3d pantograph although that may push this into a different level of funding.

 

[HerculesFerrari]

Gorton Moderne and the other changes put this over the top imo. Close enough that I'm going to paint it up and see how it goes (I'll add the federal standard paint codes assuming they work out and I'm fairly certain they will). Text line width it probably over by a hair. The mil spec calls for them not to exceed 1/8 of stroke length which means they should be dialed back closer to 0.015" at least that's an change to make.

The equator will be a bit more work since I will need a Ø4.25" steel or aluminum disc 3/32" thick with some additional features for mounting hardware. Lacking a machine shop I'll likely need to get clever. Holes were put at each end in the middle of each gimbal lock region so I can mount and remove the hemispheres with relative ease.

In the meantime I'll get started on the bezel and frame elements needed for everything to move as it should. Yep, I screwed up the 30° tick marks on this first dial . It's a 15 minute print so I'm most just happy to see everything else lining up properly.

 

[HerculesFerrari]

Had a busy weekend of working with bezel parts and just threw together a test fit with a bunch of stacked parts held together with tape and gravity. Compared to this pic from Apollo 15 I am very pleased because the spacing is very close to correct. Check out how closely they match, including parallax. Based on Smithsonian's scans it works out to 1/4" and I'm feeling pretty good with that right now.

The pointers appear to be based on mil spec ADI parts, which makes sense. I'll base the curvature and positioning off of that assumption because it's easy to change later if needed.

Some parts are deceptively simple. A great example of that is the roll pointer, which was able to simply ride along with the roll gimbal because it's the only one that's outside of the ball. The other pointers were attached to little actuators located 8-9"-ish behind the face. Everything about those is delicate because they're long, thin, have to be perfectly square, and weigh next to nothing. The actuators even have little constant force springs on either side of each needle so they will return to center.

I actually did print some rate meters but I'm questioning how those were actually set up. My understanding is that they were backlit and I noticed in several pictures that the curvature of the mask was different than the underlying backlit white parts. It's almost like the backlighting was for a different radius in an earlier version and it didn't matter if the masks needed to stand a little proud to match the pointers because it's backlit and eyes wouldn't see a difference because there weren't any shadows. That's just a guess until I can find better info.

 

[HerculesFerrari]

I've reworked several masks and the rate gauges are now all installed as modules in preparation for transillumination/backlighting. The gauge masks were printed as a trial run ahead of switching to etched aluminum. I've never tried doing acid etching with printed resin as resist so I'm hoping it works as well as it should because I'll probably be doing that a lot for this design.

 

[HerculesFerrari]

My first proper test of painting techniques has been very... Okay. The picture looks better than it does in person. I've since learned a few things about how an engraver would approach this task and have solid paint sticks on the way. Ideally that means I'll be able to spray my base coat, let it dry, rub the paint stick into all the lines, and wipe it off with a dry cloth. At least that's the theory.

For now I'll keep my expectations low and hope for the best

Other than that I'm reaching out to a couple of real FDAI owners I know to see if I can get clarification on cover glass layers and signs of transillumination. Every picture I've seen of block II CSM units looks like two layers were used. My guess is that only one was a transillumination source and the outermost cover has more to do with hermetics.

 

[HerculesFerrari]

Solid paint sticks appear to be the way. Line definition is super sharp from it being a full edge to edge fill. Now to strip, respray, and fill to verify that it's perfect.

Bonus points for the stuff smelling so much like rubbing compound.

 

[BajaTym]

Looks outstanding! Would you mind explaining how these paint sticks are used? The end results are amazing.

 

[HerculesFerrari]

Looks outstanding! Would you mind explaining how these paint sticks are used? The end results are amazing.

Thank you! I'm happy to explain and can take some pictures the next time I use them. These are the type of sticks I used. All I'm doing is rubbing them into all the lines and wiping off excess with a clean cloth. After giving it some time to dry I wet sanded very lightly with 1200 grit to remove any staining and knock residual wax/oil off the surface.

The only special thing I've found so far is that with really deep lines (these are all 0.025" iirc) you can feel when they've filled up because the sticks will stop catching on the edges.

It's normally fine to clean away excess with rubbing alcohol but I don't trust it yet for this process because I have two contrasting fill colors to deal with. It's definitely a useful technique as is but I want to try filling first and see if I can get the lines to resist paint (maybe varnish and/or wax?)

 

[BajaTym]

Thank you for relying, this is such an impressive project. The end result looks amazing. I appreciate your willingness to share/expand on your processes.

As to the paint sticks, I've seen them out-and-about, but never understood (or even considered) how they'd be used in the wild. Depending on the application and cleanup process, I can already think of endless other usabilities/possibilities. IMO, these help bring a serious measure of realism to your build. Those lines are SO clean!

 

[HerculesFerrari]

Thank you for relying, this is such an impressive project. The end result looks amazing. I appreciate your willingness to share/expand on your processes.

As to the paint sticks, I've seen them out-and-about, but never understood (or even considered) how they'd be used in the wild. Depending on the application and cleanup process, I can already think of endless other usabilities/possibilities. IMO, these help bring a serious measure of realism to your build. Those lines are SO clean!

My pleasure! There are tips and tricks all around here so I'm happy to share what I can. Here's a peek at the application process and pre-sanding gloss it creates. What I sprayed was actually matte black.

It's going to need another pass to get everything totally filled but that's fine because it's still much much faster than drop filling.