Post SOLIDWORKS Designs Here

[cwilliamrose]

I admit to having tried making one large part and then separating it into smaller pieces. I didn't like it well enough to make it my normal method. One thing is that editing the sub-part becomes an issue because everything has to be changed at the base part level. That can lead to unintended changes to the sub-parts. I much prefer to use individual parts mated together in a real-world fashion. And I don't like taking geometry from one part to drive another part's geometry either for the same reason stated above. I will sometimes start out with features that came from a mating part but I will break those associations once the design is close to being finished because the InPlace mates required to maintain your converted sketch entities is not a real-world situation. So yes, if I change one part I may need to manually change some of the mating parts too but that's where using mates works in your favor -- the mates will protest if you miss a change to one of the mated parts, kind of like real parts do when they don't fit together.

I think many SWx users would call me crazy or a dinosaur but to me working in 3D has the most value when it simulates real parts and assemblies. Since I often have to build what I design I figure it's best to make my mistakes in SWx and get it right when it's time to make chips. I'm not saying the other methods are wrong, they just don't work for me.

BTW, the race car shown in a previous post was done in one piece and then cut into sections. That is one of the few times that method paid off for me.

[Jeffrey Meyer]

Skier, I downloaded an airfoil into notepad as a txt file then opened a new part in SolidWorks. "Curve Through Reference Point" is grayed out. I can't figure out how to bring it into SolidWorks using the tuturial...I'm missing some step. Any ideas on what to try?

Hi Mark,

In a part window try Insert > Curve > Curve through XYZ Points...
You will be asked for a "Curve File", Browse for your Notepad text file and press enter.
Note that SW by default looks for a .dat file, not .txt - but when you browse for it you can change the browsing window to look for .txt file types.
Note also that the .txt file should be formatted in the form of a single line containing the delimited X,Y, & Z coordinates for each point (no headers or any other text).
The result will be a curve feature in the history tree (probably a good idea to rename the feature with the name of the airfoil).
In any sketch you can now Convert that feature into a black (fully constrained) sketch entity. By deleting the "conversion" constraint (thus making the entity unconstrained - blue) you can now do some cute manipulations. For example, in the sketch draw a straight construction line with a length equal to your required chord length and angled at your required angle of attack, and then drag the blue airfoil end points to the construction line end points. You now have your airfoil scaled and correctly oriented.

Hope this helps

[AJLiberatore]

I am not looking to share this model but several of us do share common stuff -- AN hardware, etc and the link to the library is on the Biplane Forum website. Anyway, the image below is of a Pitts project I started many years ago and will finally be putting on the front burner. This model started out as simply the aileron control system and I was looking to modify the geometry to eliminate the differential travel. I'm using different ailerons so a lot of things changed and the goal for me was to evaluate what the travels would be with stock or near-stock components and then change the bellcrank angle to get identical travels up and down. The aileron system model progressed into also looking at my elevator trim system,,,, then a full fuselage (modeled as-built since it pre-dates Solidworks) with wings. I started this model in June 2013.

I'm working with an engineer on my aileron design and we communicate via email with Solidworks models. Makes life much easier for both of us!

I have been a Solidworks user since 1998 and I see this as a tool to confirm that all the parts will play together nicely. I hope to avoid building multiple versions of parts only to find out I didn't think of all the downstream effects of my changes. The scrap pile should be smaller and pixels are much cheaper than aircraft materials. Beyond that I design the jigs and fixtures used to make some of the parts and some of those have already been built.


Attachment 5724

CW,

Your just the person I need to speak too. In my SW Class we just had a snappy introduction to the "Weldment" Module. With that in mind, I have a number of questions.

* Is that Pitts an "Assemby"?

Now onto the Fuselage, because what we have learned ( unless my teacher missed something ) is the following:

* Build your constrained wireframe ( Pitts Fuselage ) in the 3D Sketcher
* Go into the Weldment module, and draw / execute tube intersection points to get the proper fish-mouths of all tubes.
* Now from their, you have to create a model of each individual tube, save it so then it can be brought into an assembly when you are ready to do so.
* That is another and not linear step as compared to building a model(s) and bringing it into an assembly.

Is this a proper feedback? Or did my Teacher / Class miss something and the process is more like what we have learned up to know aka the model/assembly relationship?

BTW, he gave us a fictitious "Jeep Roll Cage" to do this with, I may post a snapshot later.

Tangent to this ( No pun intended ) are the longerons a spline? and how ( or ? ) are they constrained to keep them parametric?

I know this is a number of questions, however after my experience with the roll bar, I think of trying to do a simple fuselage like a Baby Ace or a Tailwind ( which might be a lot more work aka the longerons ) might be more work than a newbie might realize.

I look forward to your reply,
Anthony Liberatore

[cwilliamrose]

Hello Anthony,

I'll take a stab at your questions;

* Is that Pitts an "Assemby"?

Yes, if you mean the entire airframe. Individual parts mated together with freedom of motion where that is important (like the control system).

Now onto the Fuselage, because what we have learned ( unless my teacher missed something ) is the following:

* Build your constrained wireframe ( Pitts Fuselage ) in the 3D Sketcher
* Go into the Weldment module, and draw / execute tube intersection points to get the proper fish-mouths of all tubes.
* Now from their, you have to create a model of each individual tube, save it so then it can be brought into an assembly when you are ready to do so.
* That is another and not linear step as compared to building a model(s) and bringing it into an assembly.

No, my fuselage is a wireframe with all the tubes still in place. I see no reason to take the individual tubes and re-assemble them -- they're already assembled. Did the teacher explain what benefit doing this extra work would have?



This is the wireframe for my fuselage model.

Is this a proper feedback? Or did my Teacher / Class miss something and the process is more like what we have learned up to know aka the model/assembly relationship?

BTW, he gave us a fictitious "Jeep Roll Cage" to do this with, I may post a snapshot later.

All my extraneous stuff like fittings, tabs, etc are attached to the original weldment. Maybe there was supposed to be some learning of assemblies from this exercise but it doesn't add up to me. Maybe I'm dense and lazy, it's a strong possibility!

Here's a capture of my fuselage and the feature tree;



Notice this is a part file and you can see the wireframe sketches that define the tube locations. This is part of the fuselage assembly which adds other parts like wing and tail fittings, fin, tailwheel mount, etc.

Tangent to this ( No pun intended ) are the longerons a spline? and how ( or ? ) are they constrained to keep them parametric?

No, the longerons are not bent in gentle curves like the roll cage was probably done. A proper truss needs to have its members straight from point to point. A curved tube will buckle easily when it is in compression.

Below is the fuselage part again with the 3D sketch for the longerons highlighted;



That 3D sketch is used to create each longeron. You select the entities in that sketch to create each longeron which becomes a separate 'Structural Member' in the feature tree;


Note the arrow pointing to 3D sketch and all the trims this tube is part of.

I know this is a number of questions, however after my experience with the roll bar, I think of trying to do a simple fuselage like a Baby Ace or a Tailwind ( which might be a lot more work aka the longerons ) might be more work than a newbie might realize.

I hope the above is helpful. The weldment part is fairly simple but of you go to the VR3 website you can save some work by downloading all the tube profiles for aircraft tubing. I created many of my own before I discovered that resource.....Bill

[Jeffrey Meyer]

.....
I think many SWx users would call me crazy or a dinosaur....

Bill, your approach is by no means crazy or antiquated - on the contrary, there is a much better description for it - KIS (Keep It Simple)

While splitting parts into multiple pieces does have some advantages - like the fact that a single change in the base geometry automatically updates all the other pieces and giving you that warm feeling of achievement in your tummy, the method has a couple of major disadvantages:
1. You will probably use each piece in the form of a configuration in the multi-piece part. Then, for each change you make the software rebuilds the whole part repeatedly for each configuration. With complicated parts (read "high feature count") this process can be very compute and memory intensive (read "much thumb-twiddling").
2. If the change causes even a small geometric topology change, subsequent features can fail and result in some frustrating efforts to remove all those annoying red error triangles in the history tree.

Frankly, I prefer to design each part as a separate model and get the warm tummy at the assembly level.

There is a fairly simple way of making a compromise between the two methods:
1. Make a part representing the base geometry, and then
2. Make each of the subsequent parts separately by opening a new part and inserting the base part as the first feature.

[AJLiberatore]

Bill, Thanks.

Your Photos didn't come through in your response, or I am having a browser error. From what I can garner from your reply before seeing the photos is the Pitts Fuselage is wireframe with all tubes in place which would make it one giant "part" in the assembly, is that a correct read back? They tubes look like tubing so was the "Weldment module used, or did you Extrude-Cut to get each tube Dia / thickness and then use other Extrude-Cut to get the fish-mouths etc.?

My best,
Anthony

[cwilliamrose]

Sorry about the images. For some reason you insert the images into the body of the message and they appear again at the bottom of the post. I deleted those 'extra' images and it seemed to be OK at work. Now that I'm home the images are gone. I have no idea what the problem is... I'll get the photos back into the message tomorrow.

Yes, the fuselage structure is one big part file with all the tubes as weldment features with the intersections trimmed/extended as needed. There's a cut list and all those tubes are separate bodies. If you make a change to the fuselage geometry by changing the wireframe everything updates as you would expect. I'm not sure what Jeffery is getting at, maybe I need to be at the CAD machine to make it sink in............Bill

[Jeffrey Meyer]

I also couldn't see the images, and I'm not familiar with the Pitts fuselage structure. Can I assume that it's a fabric covered tubular structure where each piece of tubing is straight (unbent)? If so, can I assume that all the external fabric surfaces are either flat 4-sided polygons or triangular (flat by their nature)?

[cwilliamrose]

I replaced the images, hopefully everyone can see them now.

The longerons are bent but only at weld clusters, in between clusters the tube is straight.

Formers and stringers added to the structure give the fuselage a shape that is not just a box. Those stringers also support the fabric so it doesn't drum due to having large areas that are not supported. I have not defined those completely yet, only enough to create the turtledeck's shape on the sides. The bottom stringers can be seen in post #7 of this thread.

[cwilliamrose]

.....Frankly, I prefer to design each part as a separate model and get the warm tummy at the assembly level.........

The individual tubes are not created from drawings, they are fit into their locations by hand. I see no reason to make an assembly of all those tubes and suffer the resource hit the mates would require in building them into an assembly. If I change my wireframe dimensions the tubes are updated as to length and trim/extend features without red marks. There's definitely a disconnect here somewhere, I just don't see why you would want to work with anything but the original part file that contains all the tubes as weldment features. Making drawing of those tubes would be pretty much impossible anyway.

VR3 uses the tubes as individual models to machine the fishmouths on their CNC equipment. When you receive a kit from them the parts fit together like a jig saw puzzle. If you're building a fuselage from scratch you likely won't have access to the machinery VR3 has so models of the individual parts would have little or no value.