PDA

View Full Version : Modeling the Internal Structure of a Flap Assembly



Mark Meredith
01-15-2017, 04:51 PM
I've been modeling various parts of my Super Chipmunk as a way to learn SolidWorks. I'll play with a new design after I build some essential skills. I modeled the entire airplane as a single part before I realized how dumb that is, then started building more complex assemblies like landing gear to sit it on. The upside of doing this dumb thing is that I now have a fully lofted, integrated airplane with everything in the right spot and shape relative to the other parts (eg, wing, stab, fuselage all in the right relationship with each other).

The next challenge is to learn how to build from the outside-in. There is no way I'll model the whole plane - it's become clear now that I could work on it the rest of my life and never finish! I just need to do a few assemblies as learning aids. So I picked the right flap as a case study. Once the flap itself is "in hand" I'll use Layout to map out and test the whole operating system of cables, pulleys, hinges and flap handle.

Below are pics of what I have so far. It's posted on GrabCAD with the EAA handle.

To start, I split the flap from the rest of the airplane and saved it out. Then built the two end ribs from that using Convert Entities and Offset Entities to get the rib shape. The nose rib assumes .032 sheet, the tail .025, and the spar bent from .040. Goal is to use SWx Sheet Metal as much as possible, so the parts could be flattened and printed to use as cut templates.

This has been a bear to figure out, and very time consuming. I understand the popularity of hershey bar wings for RVs and Pipers - it would be SOOOO much easier to model and actually build than a tapered wing (and flap) like on a Chipmunk. The taper makes every angle different. It also means I need to figure out the shapes of each rib and model each one separately. Doing this for a whole wing (or fuselage, or stab...) using the methods I've used so far would be scary-time-consuming!

Is there a better way, or is this just the way it is? I'd appreciate insights! Things that don't work well in my model:

1. Rib flange angles are just eyeballed on the screen. They're good enough to build a form block then do the fine tuning like any builder knows how to do using fluting pliers or a shrinker machine, but they would never be good enough for SWx mates with the skin sheets (or ribs against the spar). Not that it matters that much if they mate, but it seems unprofessional to have parts that clearly don't fit perfectly.

2. Curved Flanges. Only the straight parts of the sheet metal (ribs and spar) are flanged, not any curves. I should be able to do a swept flange for the curved nose ribs but have not made SWx cooperate (even though it's easy on the Lynda training parts). Besides, they still wouldn't fit the skin because of the taper which changes the bend angle slightly from the spar to the leading edge. I experimented with making a lofted surface for the flanges (by adding a plane where the flanges end, lofting the whole flap as a surface, then inserting a new loft profile for the edges of the flanges. Then I could loft the flange surfaces and thicken them to .032 or .025.) This worked fine, and they would fit perfectly against the skin, but they're not Sheet Metal! They're just out in space, not part of the ribs, and there is no bend radius in the corners. How do designers make ribs for a tapered wing??

3. Skin Overlaps. I have not added the .032 nose skin yet. Challenge: the overlap with the tail skin at the spar. I suppose I could split then modify the Offset Entities spline at the root and tip to provide the overlap - wouldn't be hard. They could also be joggled (my Chipmunk wing is actually done this way to join the D-nose box with the rest of the wing skin aft of the spar). I've seen how to do joggles in SWx. They could also butted like on an RV wing. But the flap on the actual Chipmunk is overlapped, as are most of the fuselage panels.

4. Skin Sheet Metal. The tail skin is a lofted surface then thickened. Easy. But it's not sheet metal so can't be flattened to make a template. I know there is a way to do this with a surface. I messed with this on the complex curved parts of my fuselage - you can flatten then color code the amount of shrink and stretch to form it. It's kinda cool, and looked similar to the actual forming I had to do to make fairings on the English wheel. The flap skins are simple curves, and this should work. But is this the best way to get a skin pattern? I'd like to have the Drawing capabilites you get with SWx Sheet Metal.

That's enough challenges for now! Appreciate any suggestions. Mark :cool:
60566055

cwilliamrose
01-16-2017, 10:41 AM
Mark,

Look into two things for your ribs -- Configurations and Design Tables. I don't often use design tables but they are basically spreadsheets that drive the geometry of the configurations of the model. Editing the spread sheet edits the model. Not all dimensions are part of the spreadsheet, you select which ones are part of the design table. Being a spreadsheet means the design table can have formulas doing some heavy lifting in the background.

Design tables output configurations to the model but you can also create those manually which is what I normally do. Each dimension has a setting that will apply it all configurations or to have it be configuration specific. For your ribs you could use one rib model and make numerous configurations to change the dimensions to describe increasing larger or smaller versions of the part. If you start out with manual configurations you can then create a design table and work from the resulting spreadsheet for the rest of the versions.

I wouldn't try to make your models for the ribs so accurate that the flanges can be used as mates. I like real-world mates as much as anyone but that level of accuracy just isn't practical and it will end up making you crazy.

PM sent.......Bill

Mark Meredith
01-17-2017, 11:17 AM
I haven't tried design tables yet (as Bill suggested) to figure out rib dimensions between the root and tip on a tapered wing.
But after studying Bill's aileron and playing with different approaches to make a sheet metal flange, I came to some conclusions:

1. The Insert Bends command is a new way (to me) to add a sheet metal bend to a 3D part. Make the base shape (the rib), draw the flange on it for the planned sheet metal thickness, extrude it to make the flange, fillet the inside and outside corners so it looks like bent sheet metal, then select Insert Bends. The command converts the part to sheet metal with a flange that can be flattened. Limitations, as far as I can discern:
(A) it will only do perpendicular flanges
(B) it will only do them on a straight bend line, no curves.

2. It IS possible to form a rib flange using Swept Flange, with a curved bend line like on the top of a nose rib. It will even do them at any selected bend angle, as they would need to be on rib for a tapered wing (with acute or obtuse angle, depending on which side you add the flange). Limitations I found:
(A) SWx said the geometry is too complex when I tried to bend the flange to an inside line. In other words, the flange had to bend OUTSIDE the drawn bend line, not inside. So you'd have to figure out and allow for the bend allowance or the rib would be too small. It works, but would be hard to deal with the bend allowance instead of letting SWx figure it out.
(B) it requires circular curves, no splines. That makes it just about useless for making flanges to match an airfoil shape.

In the pic, the top flange is bent outside the drawn bend line using Swept Flange with a circular nose shape. The spar flange was added using a perpendicular extruded flange then converted to sheet metal using Insert Bends. The bottom straight flange was added using Swept Flange to the inside of the drawn bend line, at an angle. This is the only 1 of the 3 bends that would be any good on a tapered wing. This method would allow drawing the bend angle to exactly match the outside skin (whereas using Edge Flange, the angle would have to be set manually - eyeballed).

Kinda discouraging because there does not appear to be any easy solution to make a curved flange on a real airfoil, unless it's bent perpendicular. Tapered wings are great but they sure complicate matters.

6061

cwilliamrose
01-17-2017, 02:05 PM
I agree the Swept Flange tool has too many limitations to be used for ribs in a typical tapered wing. One of the biggest is that the angle of the flange is not variable. If you look at my aileron's outboard tip rib, that flange angle changes constantly -- there is no sheet metal tool in SWx that could do that part and return a flat pattern. For my purposes the model I made of that part (assuming it was fairly accurate) would be good enough to make form blocks from and that's all I really need. I could also create a model of the tooling to be CNC machined.

As for this;

>>(A) it will only do perpendicular flanges <<

You can do angles other than 90°. Sometimes it's easier to extrude the flange and the fillets together but you can make it work as separate features. As long as the fillets are concentric it will become a sheet metal feature.

Mark Meredith
01-17-2017, 06:28 PM
Ha! I got this. Here is a made-up nose rib that is modeled 100% with edge flanges. The curved and joggled flanges will bend either direction from 90 degrees. It flattens! What I discovered:

1. Curved, spline flanges will work if you start and end it with a flat line that's tangent to the curve. It only needs to be microscopically tiny, but flat.

2. Joggled flanges will work if you limit the radius and length of the flange. Limits I found: .025 sheet metal, joggle for a .032 spar, with tangent arcs of max .4 radius, with a .47 inch flange. Lengthen the flange, and the arc radius must get bigger. Bust these limits and it won't flatten. These results are not out of reason - it would be hard to actually form a rib on a bending block much tighter than this unless it's annealed.

I'll stop wearing you guys out for awhile now!
6062

cwilliamrose
01-17-2017, 06:48 PM
Thanks Mark,

I'll dig into this tomorrow. It really helps to have fresh eyes. I've been using SWx since 1998 and some things that defeated me years ago never got another look. I just found a way to create the model that avoided the limitation I knew was there.

BTW, I did play with forming tools today so I'm back up to speed on those handy little tools.........Bill