Warbler

Homebuilt Aircraft & Kit Plane Forum

Help Support Homebuilt Aircraft & Kit Plane Forum:

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Tell us about your design project, and be fully prepared for candid critiquing, which you should appreciate. Of course, you need to learn to separate the BS from the valid comments.
OK, I'll post a little here. If you'd like to talk about it, shoot a message.

I'm trying to design a light sport that fills the role of the Sonex model A. I think it's insane we aren't cutting our own kits. People don't want to build from plans anymore, and I think that's really only because the old plans are terrible compared to our standards now.

The Sonex B is great, but that fancier kit made it go outside my grad student budget. Every Sonex owner I talk to (well, at least 90%) are trying to use it as a XC flyer, and they all put autopilot in it because they all complain it's neutrally stable in pitch. It's pretty clear people are attracted to the mild aero use case but are using it for the wrong thing. I am working my way through the Chris Heintz book and putting everything in spreadsheets. I'm cross-referencing the Sonex, the Zenith 601HD, and the Zenith 650. All I want is something that stalls at 40-44 mph Vso, cruises at 120 mph, and is easy to build with minimal tooling. That 8' brake needed for most builds is an absolute project killer for the average guy in a garage. I don't know about you, but I'd much rather use the $1,800 for a blown canopy on 100LL. That, with shipping, is about a year's worth of fuel for a typical LSA pilot. Common-size channel (just like the Sonex cross-tie channel) is what I'm hoping will bring the build time down.

Most everything in LSA has those speeds already, so now it's a matter of getting the build times down. The LSA is not fancy. It's beat to death. They all look relatively the same, which gives us lots of sample data points. I accept that I will never make money in aviation, so I'm planning to take the Sandlin approach by dumping everything in the public domain, which lets me get away with using non-commercial CAD licenses.

It's not finished by any means, but here's a peek at the elementary CAD work if you want to see:

1659745166424.png

Note that the rear spar design is based on the Thatchers, and it's just a placeholder. I prefer the channel like on the Sonexes/Zeniths. There's also very little internal structure because it's hidden in the shot. I know that anybody can just dump a CAD drawing of a plane and claim wonderful things, so yes, this all means absolutely nothing yet.

-C
 

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
I have asked several questions on here about Chris Heintz's book. This is because this is one of my primary references. I wanted to show off my (incomplete) spreadsheet that follows the example at the end of the book. I've found lots of typos, mistakes, and a general lack of references. One thing this does not consider is tail loading, so that's on the to-do list. I haven't even started it yet.



You'll probably notice that in general I try to highlight negative margins of safety with a red cell. You'll also notice that I try (but am only in the middle of) to backward-engineer the 601HD as well. This gives me a good baseline of measurements. Again, none of this is really finished. It takes time.

Also, Heintz's estimates are very very conservative. This is especially the case for the spar because he assumes a uniform wing loading. This is part of the reason I'm eyeing an 18% airfoil. The preliminary runs in XFOIL/XFLR5 are showing not much of a difference between a Riblett GA30A415 and GA30A418. I'd like to take the fatter airfoil for the spar strength.
 

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
I wanted to quickly write down that the EAA has done an absolutely disgraceful job at marketing SW 3DE. I have a MS in CS, and just now learned how to access it. It's not just an entirely online environment like they marketed. THERE IS AN OFFLINE COMPONENT AS WELL. There is practically no help for this. Since Dassault advertises the online component, I figured they intended for it to be a true competitor to OnShape. My online component did not have sheet metal tools. When I reached out to Dassault, they told me I had to pay extra for the sheet metal tools. At no point did anyone tell me that there is an offline component until a guy online said, "that's odd, I have sheet metal". He didn't tell me about the offline component, so I asked, and he noted there were extra steps. It does not help that there is practically no documentation on the online component. It's mostly all marketing hype without any real information.

I encourage everyone to read this FAQ sheet from the EAA:

I'll be redrawing in SW now. Jeez, I can't believe it took me a year to get to this point.
 

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
I've been working every night trying to get caught up in SW. This is pretty exhausting after a workday as a code monkey. Truthfully, now that I've found some good extensions for OnShape (and made some quick modifications of my own), I'm still in the OS world. This project serves two purposes: (1) understand a bit about how to calculate how strong the thing is I am flying to give me some confidence when I sit in someone's plane and (2) get better at CAD. Neither of those goals are to fly an airplane I designed.

The picture attached is an example of the "modular construction" that I'm interested in exploring. I really like the channel used by the Sonex, so it looks a lot like that. The Davis DA-2 does something very similar, but the fuselage/turtledeck sides are one continuous piece that requires a form block. The goal here is to remove as many unneeded form blocks as possible to keep build time (and material waste) down.

The big takeaway from this week has been the use of the "move face" feature. I had no idea why anybody would want to use it. I have always thought that you just extrude-remove something if you need to remove material. This produces weird, unanticipated behavior for things not in-line with the coordinate system. It takes the global coordinate system into account. Move-face can remove material in-line with the part you are working on.

I have a TODO list item I'd really like to get to: write a plugin to make tabs along a curved surface. I plan to experiment with the sheet metal form plugin that someone wrote to model the wing ribs. If that doesn't work, I'll go back to drawing tabs. There are many people who don't like the individually-tabbed ribs, but they make the most sense for all-drilled at-home-CNC'ed construction. An example of someone doing this is Mr. Haines: Thatcher CX4. Doing this by hand is not great. It's not the end of the world, but it would be swell to have this done automatically.
 

Attachments

  • Screenshot (2).png
    Screenshot (2).png
    236.5 KB · Views: 0
Last edited:

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
On the topic of simplicity, I'd like for the longerons to be 1" x 1/8" angle (like the Sonex). This is a much better way to do things than the Thatcher CX4 IMHO because it cuts out any bending of longerons. Bending angle generally seems like a bad idea to me. I'd like for bent plate to be used as the gussets to connect the tail cone section to the cabin section. You can see in the other picture how I've started adding gussets for the sake of having something there.
 

Attachments

  • Screenshot from 2022-09-10 10-57-59.png
    Screenshot from 2022-09-10 10-57-59.png
    334.4 KB · Views: 1
  • Screenshot from 2022-09-10 10-57-51.png
    Screenshot from 2022-09-10 10-57-51.png
    235.8 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
A little bit of work this Saturday morning. There are some issues on Ubuntu with the graphics (thanks, NVIDIA), so I'm on Windows at the moment.

Back to the wing. Doodling some aileron stuff and cleaning up some bad references. One mistake I got used to very early on was referencing bodies instead of other sketches. This is probably the fourth time drawing a wing, and I'm finally getting the hang of the workflow. OnShape has a problem with the sheet metal tool of not always being able to move faces properly on sheet metal flanges. That's not exactly terrible, because that excess material can easily be trimmed by the builder with a set of snips. You can also see in the picture that I have the flanges not rendered so it doesn't take a minute to generate.

Screenshot (3).png

There's something of a real rear spar here now. The other chunk of what I did this morning was drawing piano hinge. Unfortunately, OnShape was not very good at this. I'll also note that the rib positions aren't finalized. I still have the numbers in crunch-limbo.

Until next time,
Connor
 

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
More doodling: Went in a new direction of modeling the skins as flat sections instead of a spline. This makes drilling holes easier with the added benefit of keeping the skin taught during drilling (straight lines are shorter from point to point). It just looks really ugly. It's pretty clear that although tabbed construction would be faster to build, there's really not a reason for it. Every other design has a solid flange, and there have been expressions of concern on this forum about straying away from the norm. That's fine. Moving on.

You can see how I doodled a baggage compartment in the wing towards the inboard section. This makes sense. Might as well use the space.

I will likely have to redraw the wing again, but I have gotten workflow down so much better than when I started. After all, getting better at CAD was actually one of the two missions of this project, so that's great. I say this because some of my references in the aileron and flap sketches are a little off, but they're still there for aesthetic reference.
 

Attachments

  • Screen Shot 2022-09-19 at 21.31.50.png
    Screen Shot 2022-09-19 at 21.31.50.png
    526.9 KB · Views: 2

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
I'd like to take a second to post some exciting updates:
1. I found a mistake in ASTM F3409-19 Standard Practice for Simplified Aircraft Loads Determination. Figures 1 and 2 were swapped. It feels good to be a part of the community. I am still waiting for them to clarify some things that didn't make sense to me.
2. I have begun moving structural calculations to a Python project, stored in a GitHub repository. The goal of this was to increase transparency. Python is pretty easy to pick up as far as programming goes, and it's fairly human-readable for people who don't code.
3. I purchased a copy of Becoming an Aircraft Stress Engineer and am a bit disappointed in the book. It does seem like a good reference for someone who does this every day and needs a cheat sheet, but for us, the Heintz book still seems the best starter choice.
 

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Bought an open-box HP Z2 Mini G5. This is one of the cheapest supported SolidWorks machines around. I had to replace the hard drive, check that the BIOS was safe, install a WLAN/BT card (seller lied on eBay), but it's running and was the same cost as my last new machine that was unreliable and got returned.

I recognize that I look really unprofessional when I bounce back and forth, but it's been a windy road to get to the point where I can run SW reliably. Finally, I can run it without it crashing every 20 minutes. This is a huge step up for me.

Now, back to drawing and learning SW.

-C
 

Attachments

  • Screenshot (1).png
    Screenshot (1).png
    517.5 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
5.1 Earthquake in Santa Clara today. This is a new one for me (although there was a small one from Southern OK that we felt in North TX a few years back). I prefer tornadoes as my natural disaster of choice. Watched a movie to chill out when I got home from work instead of going straight to SW. I got moving eventually though. Small progress tonight. I need to learn a bit about making/fixing mates, so I drew an MS20257-5 piano hinge. I don't have one in front of me, so I grabbed what I could from the Aircraft Spruce site. Close enough for now.

SW does indeed kick butt at certain things (like sheet metal flanges along a curved edge), but it certainly sucks at other things that onshape can do, like extruding a spline as a piece of sheet metal. Whoops! How do I get around this? I approximate with a series of 1" lines. Is this perfect? No! But I don't have to buy a multi-thousand-dollar piece of software or send the skins off to someone to have them flattened. Besides, we're just trying to match the precision that a typical builder who is scratch-building would achieve. The ribs will not be pre-drilled, just the skins, so we can leave the edge long and file lightly to fit if we're that worried about it.

Things that are awesome: weldments. Weldments are great for drawing the angle in the rear fuselage. This will provide a really nice cut list when done. I had originally intended to pre-drill the angle at 1.5" increments to start most of the holes. The other flange is typically not used; it's just hanging out, so the bend is generally not critical. After watching the HomebuiltHelp DVDs for the CH650, I've decided this is a bad idea. Like the ribs, the biggest bang-for-buck seems to be to pre-drill the skins. After that, the returns diminish. A confident builder can knock this out pretty fast, so it seems to me like the time here should be devoted instead to good documentation. There are lots of times when I sit around and waste time because I simply don't understand what to do.

I'd like to point out that SW has crashed once on me so far (in an empty assembly). This is a wild improvement over the last desktop I had (sad, isn't it?).
 

Attachments

  • Screenshot (5).png
    Screenshot (5).png
    371.5 KB · Views: 2

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Small progress tonight. Wasn't working hard. Drawing an aileron. Nothing really positioned in place. Drew some holes. Again, nothing really locked in. I'll start adding some planes, moving things around, and locking stuff in in a bit. Some of the flanges have to be adjusted. A lot of this is "get it down for visualization now... clean up later".
 

Attachments

  • Screenshot (7).png
    Screenshot (7).png
    545.8 KB · Views: 2

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
I've had terrible luck with top-down modeling in SW, so I've stepped back and am trying bottom-up. I haven't adjusted anything yet. Just trying to get stuff layed out a bit. Much much more to do. This isn't really much in the grand scheme of things, but I wanted to show it off. If you have advice on drawing in CAD (other than "you suck!") feel free to PM me.

-C
 

Attachments

  • Screenshot (11).png
    Screenshot (11).png
    381.9 KB · Views: 0
  • Screenshot (10).png
    Screenshot (10).png
    393 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Showing some more "hardly-works" behavior again. I can't relate flange angles to other parts in the assembly! Yuck, Stuff like this always comes as a surprise.
 

Attachments

  • Screenshot (14).png
    Screenshot (14).png
    380.7 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Still trying to figure out the whole "matching flange to part geometry". If you have advice, feel free to send it my way.
My eyes are pretty tired, but I spent a few minutes tonight goofing around and adding some angle to the rear fuselage tailcone. I don't have any extrusions drawn (because I messed up at first, and SW weldments are obnoxious to deal with).

One problem I had with sourcing aluminum angle in the past is that 3/4"x1/16" angle is too thin to consistently find. The Zodiac is a 3/4"x1/16" that mates to a bent 0.040" angle. Sonex uses 1"x1/8" in the front and rear (at least that's what I remember). Quick calculations show the 3/4" stuff is ok. I still have to do the proper calculations because things are still not finalized. You should be able to walk into a standard metal supplier and get what you need. Metal Supermarkets isn't an ideal supplier, but they carry the stuff that you're typically going to find elsewhere. Thus, good picks are 3/4"x1/8" (~2 lb/ft) or 1"x1/16" (~1.5 lb/ft and cheaper), and so the most likely choice for rear longerons is 1"x1/16".
 

Attachments

  • Screenshot (16).png
    Screenshot (16).png
    485.5 KB · Views: 2
  • Screenshot (17).png
    Screenshot (17).png
    161.3 KB · Views: 2

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Happy Thanksgiving 🦃

Doodling with a Zenith-inspired single-piece, aluminum gear hoop. I advocated against this before because of the weight and the fabrication. I have since changed my mind. I have accepted that the sonex-style of engine mount is too complicated to jig and weld. I believe it is much easier to have a tailwheel/nosewheel conversion with the aluminum gear, which is less overall work for me. The fabrication process is not as bad as I initially thought. I'm confident that a jig can be CNC-cut and used to get good, repeatable results. That's what I wanted. A press is one of those things that I can't imagine a builder would need elsewhere, which is a shame, but the mid-level 20T is available from Harbor Freight and should be enought to get the job done. I am still not thrilled about cutting 3/4" 6061-T6 (that's a lot of work), nor am I excited about the weight increase.

I moved the angle weldments from the bottom to the sides, so top and bottom rear longerons are on the same skin and can be mirrored. Drilling some holes through the fuselage angle for general ideas. Perhaps if you squint, you can see a few holes drilled in the tailcone section.

Still fighting mates. They can be a real pain. If you're really good at assemblies, please reach out. I'd really appreciate a few minutes to suggest a good workflow.

There was a teeny bit of work done on the center section. I was drawing weldments for it as well. I've read several people's posts on different SolidWorks groups advocating AGAINST using weldments, but other than Dassault's issue of making Lib Feats uneditable and placing weldments in a 'sudo' (sorry, I'm a unix person) location, they're not that bad. The real problem for me is if I need to go back and adjust them. This wasn't thought out well at all.

I spent last weekend at EAA 119 helping with Duckie, a second-hand Q2. Robert, the builder, is making great progress on it. It was nice to see his SDS installation because I've really enjoyed listening and talking to Ross. The installation is really nice. I was also slightly lured to the FlightBox system, which I've shrugged off in the past. I plan to keep in touch with Robert to see how it all fares long-term. There was also free pizza, which is always a good incentive to drive an hour and volunteer.
 

Attachments

  • Screenshot (23).png
    Screenshot (23).png
    439.9 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Time to describe some "HardlyWorks" behavior. Component Mirror operations in Assemblies do not naturally work like you think they do. There is a step 2 buried behind a tiny blue "->" button towards the top of the action menu. I couldn't for the life of me figure out why the mirrored wing was upside down. Oh well. If you don't need SW, then this is a great opportunity to mention that FreeCAD has had a recent major release that really took a step up. Unfortunately, it's still not too helpful for me.

I added an assembly to get a feel for total sizing. That wing is not the final wing. It was a lesson in how to draw a wing. It will be redrawn. I'm noticing now just how thick that 18% airfoil is in-context. Heitz used 18% on several of his designs. There's nothing wrong with it. It gets the spar tall enough to be super strong, but gosh, it does look kinda funny.

You've probably also noticed that the fuselage tailcone is too short. Again, nothing is finalized. I have to have something to look at.
 

Attachments

  • Screenshot (26).png
    Screenshot (26).png
    571.2 KB · Views: 1
  • Screenshot (27).png
    Screenshot (27).png
    788.6 KB · Views: 1
  • Screenshot (28).png
    Screenshot (28).png
    604.8 KB · Views: 1

imamac96

Well-Known Member
Joined
May 10, 2018
Messages
48
Location
North Texas / South Bay
Non-design post. Got the 35” LG curved monitor for Black Friday for $300. Wow. Money well spent. I always thought that the curbed monitors were a gimmick, but you don’t realize how much glare there is until you use one. After having one at my last job until I left about a year ago, it feels great to have one again.

Does this make me any closer to finishing? No. But gosh, it’s so much easier on my eyes to have a giant screen in front of me.

Keep building.
 

Attachments

  • 82C7A5E1-F623-4CC9-A7D6-AB253554A8ED.jpeg
    82C7A5E1-F623-4CC9-A7D6-AB253554A8ED.jpeg
    3.2 MB · Views: 1
Top