PDA

View Full Version : Modeling Tubular Frame



RCS
06-15-2017, 01:48 AM
I'm wanting to build an accurate model of the tubular frame structure of my plane - essentially reverse engineer what already exists. The Solidworks version that is available for EAA members does not seem to include the "tubing" module. My questions are: does the SW tubing module make this work substantially easier? Any tips on how to proceed without this module? I have some pretty good experience using the basic feature set, but never tackled a tubing project. Any and all help is much appreciated!

cwilliamrose
06-15-2017, 10:04 AM
Are you saying the Insert|Weldments|Structural Member... menu item and the Weldments toolbar are not there in the EAA version? That would be a huge surprise. I don't have my laptop with me but I can look at my EAA SWx install when I get home.

lathropdad
06-15-2017, 03:42 PM
The Weldment features will be there. But you have to add the tube sizes you are using.

The weldment feature is great for doing tube structures and you can make accurate templates of the mitered joints in a tube structure.

cwilliamrose
06-15-2017, 05:01 PM
You can download almost all the tubing sizes you'll ever need from VR3's website (http://vr3.ca/documents/VR3Weldments.zip). And if you follow their guidelines you can send them your model and get them to do the tube fitting for you.

RCS
06-18-2017, 01:55 PM
Thanks for pointing me in the right direction. I have never used this feature so I will need to take some time to become familiar with this. I am aware that Solidworks has a module they refer to as "Piping & Tubing" which, I believe, exposes functionality different from the weldments feature and which is not part of the EAA version of SW, but is available in one of the other academic versions. If the weldments feature will allow me to create the tubular frame accurately, then I am good with this – just need to learn a bit more about how this works. I am guessing this all starts with a good mastery of 3D sketching? Again, thanks for the help!

lathropdad
06-18-2017, 05:24 PM
I use the weldment feature frequently and found it very useful. I have modeled the Bearhawk 4 place.

You start with a 3D sketch of your fuselage. You can do a single drawing or several drawings and you can do a combination of 2D and 3D drawings With those drawings, you add/insert structural elements along the lines in the sketches. After that you trim the elements to each other. Do it right and you can do a FEA analysis of what you have drawn.

As a profession, I build tube frame formula race cars. I have done the FEA analysis and then actually tested the complete race cars and the empirical results are very close to the SW models.

The Piping and Tubing module is not what you would use for a tube fuselage.

cwilliamrose
06-18-2017, 05:58 PM
.......I am aware that Solidworks has a module they refer to as "Piping & Tubing" which, I believe, exposes functionality different from the weldments feature ........ If the weldments feature will allow me to create the tubular frame accurately, then I am good with this...... I am guessing this all starts with a good mastery of 3D sketching?....

As mentioned above, the weldment tools are what you need here, the other package is not for creating structures.

I think my Pitts fuselage (seen here) (http://eaaforums.org/showthread.php?6884-Post-SOLIDWORKS-Designs-Here&p=59663&viewfull=1#post59663) only has a few 3D sketches which are mainly for the longerons. All the other sketches are 2D. Lots of planes created for those 2D sketches but most using geometry from previous sketches for definition.

RCS
06-18-2017, 08:45 PM
This is great feedback - just exactly what I was hoping for: a starting point with some examples and helpful tips. I'm going to practice and get started. Quite sure I will have more questions - very appreciative of the support! --Rob

RCS
06-25-2017, 01:52 PM
Thanks to your help here I have built some practices models using the weldments feature - pretty easy! In my aircraft the four outer longerons are slightly curved. So in SW sketch I created a splined curve with points at each station and have them fully dimensioned (all lines are black). However I cannot add the tube to this splined line as I can do with the straight lines. Is there some step that I am missing?

cwilliamrose
06-25-2017, 03:04 PM
Tubular aircraft structures are meant to be trusses and the runs between points are almost always (and ideally) straight lines. The exception that comes to mind is the engine mount for four cylinder Lyc's. I believe your primary structure should not have any curved tubes, especially the longerons.

I'll have to take a look at my engine mount model at the shop tomorrow, that's the only weldment where I used curved tubes. It's been a few years ago since I modeled that part and I don't remember exactly what I did but I'm pretty certain the curves were done as arcs and not splines.

lathropdad
06-25-2017, 06:34 PM
I have drawn weldments may with curved tubes. Try redoing the curves with a radius or ellipse. As you are drawing, think about what you can build. I have not tried using a spline curve in a weldment and I don't know how I would bend a tube that as a spline.

RCS
06-25-2017, 07:53 PM
Yes, I do understand the engineering logic behind a truss structure, however the longerons on my Bellanca Cruisemaster are definitely curved tubing members (Guiseppe Bellanca did a lot of things his own way!). All four longerons have been formed into curved members. Now the interconnecting support members are all straight tubes. I would actually say the fuselage structure is over-built. Anyway that's my dilemma. I sincerely appreciate your help... Rob

Jeffrey Meyer
06-26-2017, 05:05 AM
I have drawn weldments may with curved tubes. Try redoing the curves with a radius or ellipse. As you are drawing, think about what you can build. I have not tried using a spline curve in a weldment and I don't know how I would bend a tube that as a spline.

I don't have too much experience with bending tubing but it seems to me the best you can achieve at home is an approximation to the curve you want. The main problem is the ends - you will always have a small straight segment at the end. But here's the thing - if you're prepared to settle for a curve that is approximated by short straight segments, then you can use any curve - circular, elliptic, conic, or spline. You can then use SWx to calculate/measure the lengths of each straight segment and the angle between segments. The shorter the segments the better the approximation, and this applies to any curve.
My 2 cents worth.

BTW - if the longerons are bent with a very large radius of curvature then it could be that the tubes are being "bent" in the elastic regime - as opposed to the plastic regime. If this is the case then the curves by their nature will be splines - not arcs/ellipses/conics, a state that SWx simulates very nicely.

cwilliamrose
06-26-2017, 07:15 AM
Jeffery,

No need to worry about making the bends perfect, once you weld the tube clusters it will all change anyway. The Bellanca used curved longerons to give the fuselage smoother contours and the tube size most likely had to be larger to provide enough strength and stiffness to make a valid structure. It's a trade-off the designer was willing to make. Most airplanes have straight tubes that approximate a curve and that method will yield a lighter structure.

Rob,

Splines aren't supported in the weldments tools. You can get close to the spline's shape using a series of tangent arcs. A bit of a PITA but you only have to do two of them (then mirror for the other two). If the spline is in 3D space instead of being on a plane it will be more work but still doable. My engine mount's curved tubes are just single arcs on planes. The curved tubes are thicker wall than the straight tubes;

6431

Jeffrey Meyer
06-27-2017, 07:37 AM
Ok Bill - Question: What tube bending equipment do you have that enables you to bend a tube to some arbitrary radius?

cwilliamrose
06-27-2017, 08:12 AM
A piece of plywood (or similar) with a radius cut on one edge with a bandsaw, then clamped to a sturdy table.

RCS
06-27-2017, 03:44 PM
Bill & Jeffrey... thank you for the ongoing dialogue - all very helpful! Just to give you the "rest of the story"... I have two working Bellanca Cruisemasters plus I also bought a fuselage from a plane that landed gear-up which I have stripped down and am using to reverse engineer various structures and components. My purpose in doing this is to make sure I know every detail of the construction of the aircraft to satisfy my engineering "need to know" obsession. From a more practical perspective I am working with the folks that have produced the Trio autopilot who have indicated that with the appropriate engineering drawings they will pursue adding the Cruisemaster to their STC list. Also designing avionics installations is much easier if the limited space behind the panel is well documented in 3D. So this is my reason for going to this effort. Again, thanks for your help!!!

cwilliamrose
06-27-2017, 05:31 PM
Interesting project. We'd love to see screen shots the fuselage model once you have something to show.

Jeffrey Meyer
06-28-2017, 01:59 PM
A piece of plywood (or similar) with a radius cut on one edge with a bandsaw, then clamped to a sturdy table.

Hi Bill - This a great idea - I'm really glad I thought of it :P;).

So, some comments on how to go about doing this:
1. IMHO the plywood mold should be made with the curve extrapolated at each end by a few cm/inches - this should be done because otherwise your bent tube will have short straight segments at each end. Of course you may be able to get away with no extrapolation of the curve if you cut the tube to be a few cm/inches longer at each end, and insert a straight bar into the extended ends.
2. Generate the curve representing the centerline of the tube in SW while taking into account that it should be a planar 2D curve. (It may be oriented in 3D, but remember that the mold/plywood is 2D). You can generate the curve using any combination of circular arcs, ellipses, conic sections, and splines.
3. Offset the curve by half the diameter of the tube.
4. Make a SW drawing of the offset curve and export it (File>Save As>*.DXF). Include a calibration line of known length in both X and Y axes.
5. Take the DXF file to your local print shop and have it printed at a 1:1 scale. Measure the actual length of the calibration lines before proceeding.
6. Stick the print to your plywood sheet using a non water based glue. (Paper expands when wet).
7. Do your stuff with the bandsaw.
8. You now have the tooling for mass production of your airframe.

cwilliamrose
06-28-2017, 02:40 PM
You forgot one important step Jeffery -- spring back. When you bend a 4130 tube around a form it will not take the shape of the form. The form has to have a tighter radius to produce the needed curve.

The short straight section at the end is not a problem, simply cut it off.

For my engine mount the curved tubes are all 3/4" OD and the bends are 12 3/8" radius to the CL of the tube so I'd start with maybe 11 3/4"R for the bending block. To lay out that radius I'd use the Omicron beam compass we have which looks like a modified tape measure, it can produce arc's as large as 72" directly on the plywood -- no need to go the the printer. After cutting my form block I'd bend a section and measure it (or compare it to a template) and adjust the form tool as needed. I'd make long enough bends to get two or three parts from each one -- I need five parts of various lengths for the mount ring.

If this was more than a simple radius, say the trailing edge tube of the rudder which has multiple tangent arcs of different radii, I'd probably make a full-size template as you describe above but it really only needs to be close since it's a weldment. I could just lay it out by hand and be plenty close enough but since I have a SWx model of the rudder getting a full-size print is the easiest way to go.

RCS
06-28-2017, 06:21 PM
I'm learning a lot more than I asked for in this thread and really pleased to read some techniques about tube bending! Back to my project for a moment... would you have a suggestion on how to orient the reference planes (datums) as I begin to build my model of the fuselage? Logically one plane will bisect the fuselage top to bottom, essentially the projection of the vertical centerline of the airframe. So considering the other two planes, maybe the other vertical plane should be establish by the mounting points of the engine mounts? What about the horizontal plane - this has me a bit stumped. Not sure what makes sense. Certainly others (probably you guys) have faced this same dilemma and hopefully have come up with a well thought out conclusion. My thanks again!!!

lathropdad
06-28-2017, 07:52 PM
Think about how you are going to build your fuselage. You will want to transfer measurements from you drawing to you building fixture or table easily.

RCS
06-28-2017, 08:33 PM
lathropdad, your point is well-taken. Although I am not building a fuselage (I am reverse engineering a certificated aircraft) I can understand your point: think ahead to consider what the model will be used for. I have some thinking to do.

cwilliamrose
06-29-2017, 07:52 AM
In the case of my fuselage I also took measurements of the actual fuselage (as welded) and modeled that instead of what the drawing said it should be when I was building it. This airplane is derived from the S-1S so it is similar in many respects but not the same. I actually started out just modeling the tail section which I did to design leading edge struts and attach points for the horizontal stab. I later used that work as the basis for the full fuselage model. I mention this because I ended up doing things bass-ackwards instead of the way the S-1S drawings were done.

So, what I did was start at the tail post and work forward instead of starting at the firewall working aft. I do have a firewall plane which gets used a lot since the front plane is back at the tailpost. Having it to do over I'd start at the firewall with the front plane. The top plane is the CL of the upper longeron tubes which are flat when viewed from the side. The right plane is on the fuselage CL as viewed from the top.

6437

6438

In my case it was easier the put the top plane on the upper longeron CL instead of at the top of the longerons as the factory drawings were done. The factory doesn't have a good way to measure from the center of a tube so they use the outside instead. When you're modeling a weldment it's easier to put your upper longeron layout sketch on the top plane so that's what I did. I do have to remember this when I'm looking at the S-1S drawings for reference.

I'm betting your fuselage isn't as simple as mine and there's no really obvious place to set the top plane. You may just have to pick something and live with it. You'll be making a lot of planes anyway so it's no big deal.

lathropdad
06-29-2017, 06:05 PM
Here is another SW weldment drawing. I did this from the fuselage drawings.

RCS
06-29-2017, 07:26 PM
That is very helpful, and actually makes good sense in my situation too. There is a tube which receives the horiz stabilizer spars that makes a perfectly logical reference feature through which the top plane of the model can pass. Thanks for the images - they help a lot.

Whallon.jesse
09-22-2017, 03:24 PM
I'm trying to follow the guidelines in this thread but I don't understand exactly how Weldments works in Solidworks. The Insert/Weldments tab is grayed out and I can't see to figure out why. Also someone above posted the dimensional drawings from VR3, can someone explain how to get those into Solidworks?

cwilliamrose
09-22-2017, 05:02 PM
The files you download from VR3 go into the SOLIDWORKS Corp\SOLIDWORKS\lang\english\weldment profiles folder. I put mine in a subfolder called 'Aircraft' so they were separate from the other non-aircraft shapes.